Tehachapi, BC

[GaryHinshaw]

Hi Ed,

I have a few of the Watchman boards from RRCirKits.  These detect current in a block, and therefore require resistor-equipped wheel sets.  I'm looking forward to hooking these up soon and trying them with the new FVM wheels.

-gfh

[GaryHinshaw]

I'll post some screen shots when I have time,

Here is a brief, virtual update.  I have been adding more of the TBC track plan into jmri's Layout Editor tool and can now report that essentially all of the functionality I was hoping to have on a control panel is readily available with the current code (and appropriate layout hardware).  Here is a screen grab that shows a portion of the layout schematic as represented in the Layout editor panel:



Key features:
* All turnouts are clickable and can be thrown from the panel.  (Not all have been decoder-equipped yet, but many have.)
* One-button routes into and out of the storage yard are enabled.  The green lights indicate the currently selected route, as do the turnout drawings.  From this diagram, the terminal manager can easily see that a route from track #7 of the storage yard is lined into track #1 (not labelled as such) of Mojave staging, and that the route is clear. 
* Red indicates occupied tracks.  (Only enabled in jmri for now - Watchman boards still need to be configured.)
* Coming soon: block detection on the upper deck mainline with virtual signal indications on the panel (and eventually on the layout...)

This panel can be easily called up on my tablet, either as a sub-panel in Engine Driver (like the screen shot a few posts up) or as a separate full-screen browser window.  So, for example, the terminal manager can run transfer jobs from any convenient location in the room.  The more I work with these tablets, the less inclined I am to mount them on the fascia as semi-permanent control panels.  It seems quite viable to have them be mobile control devices.

-gfh

[eric220]

Hi Ed,

I have a few of the Watchman boards from RRCirKits.  These detect current in a block, and therefore require resistor-equipped wheel sets.  I'm looking forward to hooking these up soon and trying them with the new FVM wheels.

-gfh

They should work very well. I've got the older RRCir-Kits tower-based occupancy detectors, and the FVM wheels work just about flawlessly.

[GaryHinshaw]

I'm very glad to here that.  I have 5 packs of FVM wheels so far, enough to equip 60 cars, a bit over 10% of my present fleet.

[Cajonpassfan]

Do you guys figure one resistor-equipped wheelset per car will do it? I have some HO friends who insist on two per car, but of course they don't have the same size fleet....
Otto K.

[eric220]

One per car should be more than sufficient. I've successfully tested with nothing more than one axel on a caboose, and nothing on intermediate cars. I'm probably going to settle for two per caboose with a one second delay on signal changes to account for momentary drops.

[Cajonpassfan]

Well, with cabooses, one would presumably have some kind of lighting anyway to generate the necessary resistance to trigger detectors? If blocks are long enough, or trains short enough, all that's needed is caboose and power resistance to trigger occupancy detectors. It's when we get into shorter OS sections, or long trains spanning blocks when intermediate wheelset resistors become necessary...and yea, one wheel set per cars would seem more than adequate.
It will be a while before I get into signaling, but I'm feeding all mainline blocks through NCE BD20 detectors while I can get to it so that I'm ready when the time comes 
Otto

[C855B]

Pretty thorough conversation about how many wheelsets per car or per train here: https://www.therailwire.net/forum/index.php?topic=36593.msg435139#msg435139

[GaryHinshaw]

Two things Otto:
1) The quintessential short block is a crossover or a turnout at a siding.  The dispatcher needs to know if a turnout is clear before throwing it.  With one wheel per car, you pretty much have that case covered (assuming you have turnouts detected separately).  Of course there can be pathological situations where, e.g., an auto rack spans a short turnout, but hey.   Oh, and jmri lets you lock turnouts that are occupied.
2) You should look into virtual signals.  Since you have block detection enabled, you can easily set up fully capable virtual signals in jmri with logic derived from the detection sensors.  This approach lets you test and modify your signalling scheme before committing to a specific hardware configuration, and it gets you up and running with signals much sooner than you otherwise would.

FVM resistor wheels keep showing up in small batches at MBK.  I just ordred 3 more packs tonight... and they're sold out again.   

[C855B]

... Of course there can be pathological situations where, e.g., an auto rack spans a short turnout ...

Be thinking two racks or other 86' footers coupled together with the resistors at opposite ends. That's ~170' feet of undetectable, which will certainly play havoc with in-turnout detection. So therefore the "rule" is one resistor wheelset per one-half of the length of the shortest detection block. Some of this can be mitigated with time delays on the detection release, but that wouldn't account for stopped trains.

However, it just dawned on me - my old saw that "detection is detection". Turnout occupancy can use optical detection. That significantly reduces the demand to have a "perfect" fleet of resistor-equipped cars.

[Cajonpassfan]

Thanks for the advice guys, and yes I'm ordering more wheelsets every chance I get when they are available.

But I'm afraid signalling my layout is a much simpler proposition, because it's based on an actual prototype in the days before CTC and dispatcher controlled turnouts. Except for a few locations at towers and such, switches were/are lined by traincrews. And, for the most part, signaling on the doubletrack was/will be one-directional, just like the real thing.
Given my setting, there is no need for short OS sections such as over a turnout, so I'm wondering whether equipping every car with resistors is overkill..... but I can see a local leaving a few cars without a caboose on the main while switching and that would require resistored axles....

Nevertheless, I've operated on fully signalled JMRI based/CTC run layouts, and it is a blast!
Regards, Otto

[wcfn100]

Two things Otto:
1) The quintessential short block is a crossover or a turnout at a siding.  The dispatcher needs to know if a turnout is clear before throwing it.  With one wheel per car, you pretty much have that case covered (assuming you have turnouts detected separately).  Of course there can be pathological situations where, e.g., an auto rack spans a short turnout, but hey.   Oh, and jmri lets you lock turnouts that are occupied.

Why can't you just make the turnout blocks a little longer?

Jason

[jagged ben]

What it comes down to is this...

If you want to rely on detected wheelsets to ensure that a car is not fouling a turnout, then your rail gap needs to be farther from the fouling point than the end of the car is from the wheelset.

Keep reading that til you get it.

If you only have one detector wheelset per car then the distance in question is going to be significant on some cars.  e.g. nearly 6 inches instead of 1 inch on a 89ft flat or autorack.   (A 40ft boxcar, by comparison, is only about 2.25" vs .75".) 

My suggestion would be to pick a standard that suits your fancy, and put two wheelsets on cars that are longer than that standard.   For example, you could say "My standard is anything up to a 60ft boxcar only gets one wheelset."  Okay,  it's about 4inches from the inside wheelset to the end of the car on a 60ft boxcar, so put your gaps 4 inches back from the actual fouling point of the turnout.    Longer cars, like 89 footers, get two wheelsets.   

The disadvantage of choosing a longer distance like this is that a train which actually fits inside the fouling points of a siding may not fit electrically if the wheelsets happen to be on the wrong end of the car.  It won't happen often, but it maybe sometimes.

Alternatively, you could say "I don't ever want to waste an inch of siding."   Okay, put two wheelsets on every car, and make your gaps 1" from the fouling points.  Your sidings just all got 6" longer.

Me personally, I'd probably do two wheelsets on anything longer than a 50ft boxcar.
 

[Santa Fe Guy]

Otto. On my SFRSD we thought that one in three cars should only have a resistor equipped wheel set. Nope that didn't work so all cars are now fitted with one axle (FVM) equipped with a resistor fitted by my young well sighted friend.
Also when we set up for CTC, I included the turnouts in the block, wrong, the dispatchers panel could not tell if the train had or had not cleared the turnout so all turnouts are now individually detected and it is so much better.
Regards.
Rod.

[C855B]

Hey, Gary... have you seen the news about the mudslides in California? The big one that swallowed 200+ vehicles was at Sand Canyon Rd., just a couple of miles east of Monolith. I haven't heard anything yet about how the RR fared, but I'll betcha it's a huge mess since (as you know) the line is right next to the wash.

EDIT: The report is the track did more or less OK. The mudslide was on the Hwy 58 side, and the freeway took the worst of it. Track was covered but not moved and they stopped trains for most of the day while they cleared the muck. Word is operations were resumed late Friday evening.